This project will investigate why languages combine sounds in some ways and not in others. Every language shows preferred patterns for combining the particular sounds which form its phonemic inventory and rules out many other patterns. Comparisons across languages show that many of these patterns are similar from language to language and recur in many unrelated languages. When widely repeated similarities are found across languages it is most likely because of properties built into the human language capability. Cross-language studies that are sufficiently broad in scope provide a means by which such properties governing the design of languages can be studied. This project will be based on a study of the patterns of permitted sound sequences in a carefully structured sample of languages and will present a picture of the language-universal elements in these patterns. However, the patterns themselves are not the primary objects of research, but rather the principles of language design which produce them. The main hypothesis is that a basic principle at work is the optimization of a trade-off between simplicity in articulatory sequencing and distinctiveness in acoustic sequencing. While this hypothesis has been put forward in the past, its validity and range of application have not been rigorously explored. It is predicted that preferences for certain syllable structures can be subsumed under the same principles which govern combinations of individual sounds. The trade-off hypothesis will be directly addressed by careful analysis of the survey data. Findings will also be evaluated in terms of a model of articulatory/acoustic relations being developed at UCLA. This will allow such clinically useful notions are articulatory difficulty to be expressed in quantitative terms. The size and structure of the units over which sequencing patterns apply will also be determined. These units reflect the size and structure of units in the mental representation of language and in the motor programming of speech production. This research will thus probe fundamental components of the human language faculty and will suggest better ways of modeling these components in terms relevant to both general and clinical linguistics.